paper proposes an integrated method for efficient content based image
retrieval using color, shape and rotational invariant texture features.
The present paper derived rotational invariant features on each region.
To derive shape features textons are computed. To represent texture
features gray level co-occurrence matrix (GLCM) features are derived on
region based rotational invariant texton matrix. These features are
combined with HSV histograms. The advantage of region based models is
they are more applicable when working
with images of large size and especially in real time environment. The
image retrieval is performed on five categories of Wang database and the
present method is compared with texton co-occurrence matrix (TCM),
color correlogram gradient (CCG) and GLCM methods
Uploaded by ICGST LLC on